Computational approaches to the inference of biological network connectivity
The phenotype of a unicellular organism is determined by an integrated network of genes, proteins, and metabolites that participate in reciprocal regulatory relationships. Creating a quantitative description of this network is essential to understanding, predicting and manipulating cellular behavior. A first step toward this goal is deciphering the connectivity of the network, i.e., the pattern of interactions among its components, and its general organizational features. Given the complexity of this task, it is convenient to view the integrated network as a group of superimposed subnetworks, including the gene regulatory, protein, and metabolic networks. This thesis addresses the inference of connectivity in the gene and protein networks of Saccharomyces cerevisiae and Pseudomonas aeruginosa.
It is widely believed that biological networks are organized into functional modules, defined here as groups of genes, proteins, and other small molecules that participate in common subcellular processes. Chapter 2 presents a computational method to identify transcriptional coordination among predefined functional modules of genes in S. cerevisiae. Two modules are said to be coordinated by a particular transcription factor if it confers a distinct expression profile on its target genes in each module. The method was applied to a variety of module pairs to reveal a global network of functional coordination in yeast.
Chapter 3 describes an algorithm that searches the S. cerevisiae protein-protein interaction network for signal transduction pathways whose components are encoded by coexpressed genes. This enables the discovery of known and novel linear signal transduction pathways. When linear pathways with common endpoints are combined, the resulting network represents the complex interactions underlying signal transduction more fully than a traditional linear representation.
Chapter 4 describes the identification of putative transcription factor binding motifs that shape the topology of the quorum sensing gene regulatory network in the pathogenic bacterium P. aeruginosa. The approach identified novel, putative motifs, and variants of a motif known to function in quorum sensing. These motifs may imbue quorum-sensing regulation with signal and/or temporal specificity.